Initial grain dimension on texture, microstructure and kinetics

There are a number of aspects involving the deformed orientations may contribute different recrystallization behaviours in a material. These aspects are independent of material chemistry but evolve because of deformation orientation-disorientation and strain energy distribution in deformed orientations. Each initial grain has a particular orientation and an average dislocation density in this model. Experimental evidences of grain dependent deformation energy and accumulated long and short range disorientation were obtained from the substructure study of aluminium with similarly oriented grains with different diameter (200 micrometer and 400 micrometer) (Fig. 42).

The substructure and disorientation development in smaller grain (200 micrometer) material is more pronounced than those of larger grain (400 micrometer) material. The effect of initial grain dimension dependent deformation energy and disorientation distribution goes in the recrystallization simulations through the combination of GIA and 3IVM model.

(a) (b)

(c) (d)

(e)

(f)

0 10 20 30 40 50 60

0 100 200 300 400

Lenth Scale(micron)

Long Range Misorientation

200_Grain1 200_Grain2 200_Grain3 200_Grain4 200_Grain5 200_Grain6

0 5 10 15 20 25

0 50 100 150 200 250

Length Scale (micron)

Long Range Misorientation

400_Grain1 400_Grain2 400_Grain3 LongTranverse Short Transverse 45°toLongTransverse

Fig. 42(a) Texture of aluminium with around 400-micrometer grain diameter, type 1. (b) Texture of aluminium with around 200-micrometer grain diameter, type2. (c) Long range disorientation for type 1

material. (d) Long range disorientation for type 2 material. (e) 60% cold rolled (true strain 0.8) type 1 material, (f) 60% cold rolled (true strain 0.8) type 2 material. (Courtesy: Prof. I Samajdar, S.

Badirujjaman).

The deformed orientations with respective dislocation densities were randomly distributed in form of grain aggregates according to the GIA 8 grain cluster concept and necessary precautions were taken to restore the deformation texture. Recrystallization simulations were performed varying the initial grain dimension and keeping the nuclei

density constant for a particular distribution of deformed orientations to know the effect of initial grain dimension, the location of nucleation and the orientation of nucleation.

The simulation parameters involving the nuclei locations and orientations are given below.

(i) Nuclei locations were random and orientations were randomly selected.

(ii) Nuclei locations were along grain boundary and orientations were randomly selected.

(iii) Nuclei locations were along grain boundary and orientations of original grains were selected.

Since texture orientations are randomly selected in simulation (i) and in simulation (ii), the difference in results should be created by the location of nuclei in terms of their surroundings.

The results from simulation (ii) and from simulation (iii) should clarify the effect of nuclei orientations where locations of nuclei in both simulations are along grainboundary.

In aluminium alloys there are possibility of nuclei orientation change depending on pre-recrystallization deformation. A very high degree pre-pre-recrystallization deformation may randomise nuclei orientations otherwise the grain orientation should remain preserved in nucleation. Moreover, in a 2 phase aluminium alloy where majority of nuclei produced from precipitates the number of nuclei or new grains should not vary to a large extent with the pre-recrystallization deformation.

The following experimental work on aluminium alloys was performed to confirm the earlier mentioned possibilities. The alloy was recrystallized in salt bath furnace after pre-recrystallization 60%, 80% and 90% cold rolling. Necessary precautions were taken to avoid grain growth after recrystallization. The annealing operation were carried out in three different temperature levels i.e. 300°C, 350°C and 400°C. The microstructrure of annealed samples revealed that samples were fully recrystallized after 3.5 minutes of annealing at 400°C and microstrucrtures gain stability after that because of pinning of grain boundaries by precipitates. The results of 400°C annealed material show insignificant variation in total number of grains, maximum, minimum and in average grain dimension for all samples with different levels of pre-recrystallization deformations (Fig. 43). However, differences in recrystallization grain distributions and in recrystallization textures are very much clear for all three samples with different pre-recrystallization deformations. These findings are in favour of the fact that the degree of

pre-recrystallization cold rolling can change nuclei and recrystallization textures but it has less effect on the total number of recrystallization grains for an alloy which contains a large number of precipitates capable of stimulating recrystallization nucleation (Fig.

43 - Fig. 45).

A similar condition was tried to simulate in simulation (ii) and in simulation (iii) where the nuclei orientations were varied keeping nuclei locations and number constant.

a) b)

c)

0 5 10 15 20 25

60%

CR_AVG 80%CR_AVG

90%

CR_AV G

60%

CR_MIN 80%

CR_MI N

90%

CR_MI N

60%

CR_MAX 80%

CR_MAX 90%

CR_MAX 300°C 350°C 400°C

MIN: MINIMUM, MAX: MAXIMUM, CR:

COLD ROLLED, AVG: AVERAGE (d)

Fig. 43 a) 60% cold rolled (true strain 0.8) and annealed (Salt bath 400°C), b) 80% cold rolled (true strain 1.6) and annealed (Salt bath 400°C), c) 90% Cold Rolled (true strain 2.4) cold rolled and annealed (Saltbath 400°C), d) Recrystallization grain size with different prior recrystallization cold rolling at

different saltbath temperatures.

-5

Fig. 44 Grain distribution in recrystallization microstructures obtained from salt bath annealing at 400°C for three differently cold rolled (60%, 80% and 90%) sample prior to the annealing.

0

Fig. 45 Variation in volume fraction of different recrystallization texture components with prior recrystallization cold rolling (true strain 0.8, 1.6 and 2.4).

The results listed below obtained from varying the initial grain dimension where both nuclei locations and the nuclei orientations are random. The intensity of rolling texture components are found to reduce after recrystallization with the increase in the initial grain dimension. The increase in frequency of larger recrystallized grains with the increase in initial grain volume creates the textural change. The slight increase in near {100}<100> orientation and a faster recrystallization kinetics are associated with the increase in initial grain volume (Fig. 46). The comparison of recrystallization texture results from simulations (i) and (ii) (Fig. 46 and Fig. 47) shows no major variation which indicates a negligible effect of nuclei locations when the nuclei orientations are randomly selected.

Fig. 46 a) The effect of initial grain dimension (Left to Right: (25-40-60) micron) on simulated textures.

C: Cube G: Goss Co: Copper R: R component B: Brass

0 2

%

4 6

olu

8 10

eFr

io 12

n 14

C G Co R B

Texture Com ponents

Vmact

16

25micron 40micron 60micron

Fig. 46 b) The effect of initial grain dimension on volume fraction of simulated texture components.

Initial grain dimension

25micron

Initial grain dimension

40micron

Initial grain dimension 60micron Fig. 46 c) The simulated recrystallized microstructure from initial grains of different dimensions.

0

Fig. 46 d) The calculated recrystallized grain distribution with different initial grain dimensions. Grains are in cell numbers.

0

Fig. 46 e) The calculated Recrystallization kinetics and recrystallization exponents with different initial grain dimensions.

The variation in results obtained from two types of grain boundary nucleations in simulation (ii) and in (iii) is note worthy. The random selection of nuclei orientations generates weak texture. The intensity of texture increases sharply when the original grain orientation is selected (Fig. 47 b,d)). The major consequence of grain oriented nuclei selection is the significant reduction in intensity of Cu texture orientation after recrystallization. The previously found effect of initial grain dimension in recrystallization texture and grain distribution from simulation (i) is preserved in those of simulation (ii) and in simulation (iii). The random selection of nuclei orientations improves the recrystallization kinetics, which is faster than the recrystallization kinetics obtained from grain oriented nuclei (Fig. 51).

(a) (b)

(c) (d)

Fig. 47 The effect of nuclei orientation and grain dimension on the simulated recrystallization texture.

a) 25-micrometer initial grain diameter. Nuclei orientation randomly selected, b) 25-micrometer initial grain diameter. The grain orientation is selected as nuclei orientation. c) 60-micrometer initial grain

diameter. Nuclei orientation randomly selected. d) 60-micrometer initial grain diameter. The grain orientation is nuclei orientation.

0 5 10 15 20 25

cube goss

copper copper2

S/R S/R2

S/R3 S/R4

brass brass2 Te x ture Compone nts

Texture Volume Fraction

25micron_Ori 25micron_Ran 60micron_Ori 60micron_Ran

Fig. 48 The simulated volume fractions of texture components from different nuclei orientations (Ori:

initial grain orientation and Ran: random orientation) and different initial grain dimensions (25 micron and 60 micron).

Fig. 49 a) The effect of nuclei orientation and initial grain size on simulated recrystallization microstructure. 25-micrometer initial grain size.

Nuclei orientation randomly selected.

Fig. 49 b) The effect of nuclei orientation and initial grain size on simulated recrystallization microstructure. 25-micrometer initial grain size.

Nuclei orientation is grain orientation.

Fig. 50 a) The effect of nuclei orientation and initial grain size on simulated recrystallization microstructure. 60-micrometer initial grain size.

Nuclei orientation randomly selected.

Fig. 50 b) The effect of nuclei orientation and initial grain size on simulated recrystallization microstructure. 60-micrometer initial grain size.

Nuclei orientation is grain orientation.

0

Fig. 51 a) The simulated recrystallization kinetics with variable initial grain diameter and nuclei

orientation (random and grain orientation).

-5

Fig. 51 b) The simulated recrystallization exponents with variable initial grain diameter and nuclei orientation (random and grain orientation).

The effect of change in intensity of oriented growth was studied with the 60 micron size initial grains, constant number of grain oriented nuclei. Though the texture intensity reduces compared to that of Fig. 47 d), The volume fraction near {100}<100> grains

increases with the increase in degree of oriented growth. The frequency of larger grains also increases (Fig. 52).

Fig. 52 Simulated recrystallization microstructure and texture with improved oriented growth.